You are in:
Tagging helps improve incident reconstruction
Continuing with our ‘Did You Know…’ series, this month Murray Barker, product manager public safety, NICE Systems provides a view on how tagging calls improves incident reconstruction capability and the overall performance of the emergency service control room…
Tagging audio interactions between emergency control room call takers and members of the public can improve the process of incident reconstruction, identify specific trends in crime, aid investigations and also develop the overall performance and response of the control room.
Most emergency service control rooms are now recording and storing all audio interactions, in order to improve incident management, scenario reconstruction and response time. These improvements are of course dependant on the method in which information is extracted from these interactions. Tagging a call for a specific reason is an effective way of enabling the identification of a call, or a group of calls, that may contain specific information, which can be searched for and retrieved in a timeframe that improves incident reconstruction capability.
Each call the control room receives holds within it specific information which, under further analysis, may be a vital piece of evidence or a crucial link in scenario reconstruction. Typically, once calls have been recorded, a control room supervisor will manually annotate the audio or video recordings according to the content and outcome of each interaction. These annotations can be made through text, voice or book marking (a method of flagging up a call with a general alert).
Search and retrieve
The information annotated on the tagged interactions enables the search, identification and immediate retrieval of specific information. For example, searching for annotated words such as ‘gun crime’ or ‘domestic violence’ will provide the control room supervisor with calls that have previously been annotated using these words.
Such annotations can be used for specific purposes, for example, the search and retrieval of further information on an individual incident. If an investigation into a gun crime leads authorities to believe that the suspect may have previously offended, a search for the right annotation content will flag up the relevant stored interactions.
Manual annotation can also be used to evaluate specific trends in crime, enabling authorities to monitor and find new patterns. Searching tagged interactions that are annotated with words such as ‘car crime’ can provide a user with an understanding of the trends and patterns of the crime, identifying details such as the geographical locations and times the incidents have occurred. This information can then be used to pinpoint specific patterns, offering further insight into an investigation.
As well as enhancing incident management and identifying trends in specific crimes, manual annotation may also improve the overall performance of the emergency control room. With the HMIC (Her Majesties Inspectorate of Constabulary) regulating call handling standards and monitoring control room targets, quality management is of paramount importance for the emergency control room.
Improving quality management
Using the interactions of the control room by manually annotating the outcome of a call is also a method of tagging that can improve quality management. Manual annotation allows the interaction between the call taker and member of the public to be tagged according to the performance of each call taker. As a result, a control room supervisor can analyse interactions in specific circumstances and recognise and address both positive and negative responses. The tagged calls become indicative of how both call takers and members of the public react in certain situations. A search on all annotations where specific emergencies such as ‘domestic violence’ have occurred would enable the control room supervisor to identify how call takers have responded to domestic violence related calls, giving an insight into the best possible practice.
With the advances in audio analytic techniques such as word spotting, emotion detection and speech analysis, it is now possible to monitor ‘all’ control room interactions, specifically tagging them as they occur. Analytic techniques allow automatic recognition of specific words or voice stress, and can also alert a user in near-real time. This enables call content to be identified automatically, therefore interactions do not have to be reviewed before being categorised. Ultimately, this means that the use of analytic techniques can provide both a proactive and retrospective solution.
Combining these audio analytic techniques with manual annotation would further enhance the capability of retrieving information. Manual annotation allows stored interactions to be identified in terms of the context found within the interaction but, combined with the capability of techniques such as real time automatic recognition and emotion detection, the search, identification and retrieval process advances to a higher level.
